Comparative Analysis of ARIMA and LSTM Model-Based Anomaly Detection for Unannotated Structural Health Monitoring Data in an Immersed Tunnel

Structural Health Monitoring(SHM)systems have become a crucial tool for the operational management of long tunnels.For immersed tunnels exposed to both traffic loads and the effects of the marine environment,efficiently identifying abnormal conditions from the extensive unannotated SHM data presents a significant challenge.This study proposed amodel-based approach for anomaly detection and conducted validation and comparative analysis of two distinct temporal predictive models using SHM data from a real immersed tunnel.Firstly,a dynamic predictive model-based anomaly detectionmethod is proposed,which utilizes a rolling time window for modeling to achieve dynamic prediction.Leveraging the assumption of temporal data similarity,an interval prediction value deviation was employed to determine the abnormality of the data.Subsequently,dynamic predictive models were constructed based on the Autoregressive Integrated Moving Average(ARIMA)and Long Short-Term Memory(LSTM)models.The hyperparameters of these models were optimized and selected using monitoring data from the immersed tunnel,yielding viable static and dynamic predictive models.Finally,the models were applied within the same segment of SHM data,to validate the effectiveness of the anomaly detection approach based on dynamic predictive modeling.A detailed comparative analysis discusses the discrepancies in temporal anomaly detection between the ARIMA-and LSTM-based models.The results demonstrated that the dynamic predictive modelbased anomaly detection approach was effective for dealing with unannotated SHM data.In a comparison between ARIMA and LSTM,it was found that ARIMA demonstrated higher modeling efficiency,rendering it suitable for short-term predictions.In contrast,the LSTM model exhibited greater capacity to capture long-term performance trends and enhanced early warning capabilities,thereby resulting in superior overall performance.

An improved typhoon monitoring model based on precipitable water vapor and pressure

The potential of monitoring the movement of typhoons using the precipitable water vapor(PWV) has been confirmed. However, monitoring the movement of typhoon is focused on PWV, making it difficult to describe the movement of a typhoon in detail minutely and resulting in insufficient accuracy. Hence,based on PWV and meteorological data, we propose an improved typhoon monitoring mode. First, the European Centre for Medium-Range Weather Forecasts Reanalysis 5-derived PWV(ERA5-PWV) and the Global Navigation Satellite System-derived PWV(GNSS-PWV) were compared with the reference radiosonde PWV(RS-PWV). Then, using the PWV and atmospheric parameters derived from ERA5, we discussed the anomalous variations of PWV, pressure(P), precipitation, and wind speed during different typhoons. Finally, we compiled a list of critical factors related to typhoon movement, PWV and P. We developed an improved multi-factor typhoon monitoring mode(IMTM) with different models(i.e.,IMTM-I and IMTM-II) in different cases with a higher density of GNSS observation or only Numerical Weather Prediction(NWP) data. The IMTM was evaluated through the reference movement speeds of HATO and Mangkhut from the China Meteorological Observatory Typhoon Network(CMOTN). The results show that the root mean square(RMS) of the IMTM-I is 1.26 km/h based on ERA5-P and ERA5-PWV,and the absolute bias values are mostly within 2 km/h. Compared with the models considering the single factor ERA5-P/ERA5-PWV, the RMS of the IMTM-I is improved by 26.3% and 38.5%, respectively. The IMTM-II model manifests a residual of only 0.35 km/h. Compared with the single-factor model based on GNSS-PWV/P, the residual of the IMTM-II model is reduced by 90.8% and 84.1%, respectively. These results propose that the typhoon movement monitoring approach combining PWV and P has evident advantages over the single-factor model and is expected to supplement traditional typhoon monitoring.

Residual subsidence time series model in mountain area caused by underground mining based on GNSS online monitoring

The residual subsidence caused by underground mining in mountain area has a long subsidence duration time and great potential harm,which seriously threatens the safety of people's production and life in the mining area.Therefore,it is necessary to use appropriate monitoring methods and mathematical models to effectively monitor and predict the residual subsidence caused by underground mining.Compared with traditional level survey and InSAR(Interferometric Synthetic Aperture Radar)technology,GNSS(Global Navigation Satellite System)online monitoring technology has the advantages of long-term monitoring,high precision and more flexible monitoring methods.The empirical equation method of residual subsidence in mining subsidence is effectively combined with the rock creep equation,which can not only describe the residual subsidence process from the mechanism,but also predict the residual subsidence.Therefore,based on GNSS online monitoring technology,combined with the mining subsidence model of mountain area and adding the correlation coefficient of the compaction degree of caving broken rock and the Kelvin model of rock mechanics,this paper constructs the residual subsidence time series model of arbitrary point on the ground in mountain area.Through the example,the predicted results of the model in the inversion parameter phase and the dynamic prediction phase are compared with the measured data sequence.The results show that the model can carry out effective numerical calculation according to the GNSS monitoring data of any point on the ground,and the model prediction effect is good,which provides a new method for the prediction of residual subsidence in mountain mining.

Big Model Strategy for Bridge Structural Health Monitoring Based on Data-Driven, Adaptive Method and Convolutional Neural Network (CNN) Group

This study introduces an innovative“Big Model”strategy to enhance Bridge Structural Health Monitoring(SHM)using a Convolutional Neural Network(CNN),time-frequency analysis,and fine element analysis.Leveraging ensemble methods,collaborative learning,and distributed computing,the approach effectively manages the complexity and scale of large-scale bridge data.The CNN employs transfer learning,fine-tuning,and continuous monitoring to optimize models for adaptive and accurate structural health assessments,focusing on extracting meaningful features through time-frequency analysis.By integrating Finite Element Analysis,time-frequency analysis,and CNNs,the strategy provides a comprehensive understanding of bridge health.Utilizing diverse sensor data,sophisticated feature extraction,and advanced CNN architecture,the model is optimized through rigorous preprocessing and hyperparameter tuning.This approach significantly enhances the ability to make accurate predictions,monitor structural health,and support proactive maintenance practices,thereby ensuring the safety and longevity of critical infrastructure.

Contribution of the MERISE-Type Conceptual Data Model to the Construction of Monitoring and Evaluation Indicators of the Effectiveness of Training in Relation to the Needs of the Labor Market in the Republic of Congo

This study proposes the use of the MERISE conceptual data model to create indicators for monitoring and evaluating the effectiveness of vocational training in the Republic of Congo. The importance of MERISE for structuring and analyzing data is underlined, as it enables the measurement of the adequacy between training and the needs of the labor market. The innovation of the study lies in the adaptation of the MERISE model to the local context, the development of innovative indicators, and the integration of a participatory approach including all relevant stakeholders. Contextual adaptation and local innovation: The study suggests adapting MERISE to the specific context of the Republic of Congo, considering the local particularities of the labor market. Development of innovative indicators and new measurement tools: It proposes creating indicators to assess skills matching and employer satisfaction, which are crucial for evaluating the effectiveness of vocational training. Participatory approach and inclusion of stakeholders: The study emphasizes actively involving training centers, employers, and recruitment agencies in the evaluation process. This participatory approach ensures that the perspectives of all stakeholders are considered, leading to more relevant and practical outcomes. Using the MERISE model allows for: • Rigorous data structuring, organization, and standardization: Clearly defining entities and relationships facilitates data organization and standardization, crucial for effective data analysis. • Facilitation of monitoring, analysis, and relevant indicators: Developing both quantitative and qualitative indicators helps measure the effectiveness of training in relation to the labor market, allowing for a comprehensive evaluation. • Improved communication and common language: By providing a common language for different stakeholders, MERISE enhances communication and collaboration, ensuring that all parties have a shared understanding. The study’s approach and contribution to existing research lie in: • Structured theoretical and practical framework and holistic approach: The study offers a structured framework for data collection and analysis, covering both quantitative and qualitative aspects, thus providing a comprehensive view of the training system. • Reproducible methodology and international comparison: The proposed methodology can be replicated in other contexts, facilitating international comparison and the adoption of best practices. • Extension of knowledge and new perspective: By integrating a participatory approach and developing indicators adapted to local needs, the study extends existing research and offers new perspectives on vocational training evaluation.

Temporally Preserving Latent Variable Models:Offline and Online Training for Reconstruction and Interpretation of Fault Data for Gearbox Condition Monitoring

Latent variable models can effectively determine the condition of essential rotating machinery without needing labeled data.These models analyze vibration data via an unsupervised learning strategy.Temporal preservation is necessary to obtain an informative latent manifold for the fault diagnosis task.In a temporalpreserving context,two approaches exist to develop a condition-monitoring methodology:offline and online.For latent variable models,the available training modes are not different.While many traditional methods use offline training,online training can dynamically adjust the latent manifold,possibly leading to better fault signature extraction from the vibration data.This study explores online training using temporal-preserving latent variable models.Within online training,there are two main methods:one focuses on reconstructing data and the other on interpreting the data components.Both are considered to evaluate how they diagnose faults over time.Using two experimental datasets,the study confirms that models from both training modes can detect changes in machinery health and identify faults even under varying conditions.Importantly,the complementarity of offline and online models is emphasized,reassuring their versatility in fault diagnostics.Understanding the implications of the training approach and the available model formulations is crucial for further research in latent variable modelbased fault diagnostics.

Genetic Regression Model for Dam Safety Monitoring

Under-fitting problems usually occur in regression models for dam safety monitoring.To overcome the local convergence of the regression, a genetic algorithm (GA) was proposed using a real parameter coding, a ranking selection operator, an arithmetical crossover operator and a uniform mutation operator, and calculated the least-square error of the observed and computed values as its fitness function. The elitist strategy was used to improve the speed of the convergence. After that, the modified genetic algorithm was applied to reassess the coefficients of the regression model and a genetic regression model was set up. As an example, a slotted gravity dam in the Northeast of China was introduced. The computational results show that the genetic regression model can solve the under-fitting problems perfectly.

Working Condition Real-Time Monitoring Model of Lithium Ion Batteries Based on Distributed Parameter System and Single Particle Model

Lithium ion batteries are complicated distributed parameter systems that can be described preferably by partial differential equations and a field theory. To reduce the solution difficulty and the calculation amount, if a distributed parameter system is described by ordinary differential equations （ODE） during the analysis and the design of distributed parameter system, the reliability of the system description will be reduced, and the systemic errors will be introduced. Studies on working condition real-time monitoring can improve the security because the rechargeable LIBs are widely used in many electronic systems and electromechanical equipment. Single particle model （SPM） is the simplification of LIB under some approximations, and can estimate the working parameters of a LIB at the faster simulation speed. A LIB modelling algorithm based on PDEs and SPM is proposed to monitor the working condition of LIBs in real time. Although the lithium ion concentration is an unmeasurable distributed parameter in the anode of LIB, the working condition monitoring model can track the real time lithium ion concentration in the anode of LIB, and calculate the residual which is the difference between the ideal data and the measured data. A fault alarm can be triggered when the residual is beyond the preset threshold. A simulation example verifies that the effectiveness and the accuracy of the working condition real-time monitoring model of LIB based on PDEs and SPM.

A landslide monitoring method using data from unmanned aerial vehicle and terrestrial laser scanning with insufficient and inaccurate ground control points

Non-contact remote sensing techniques,such as terrestrial laser scanning(TLS)and unmanned aerial vehicle(UAV)photogrammetry,have been globally applied for landslide monitoring in high and steep mountainous areas.These techniques acquire terrain data and enable ground deformation monitoring.However,practical application of these technologies still faces many difficulties due to complex terrain,limited access and dense vegetation.For instance,monitoring high and steep slopes can obstruct the TLS sightline,and the accuracy of the UAV model may be compromised by absence of ground control points(GCPs).This paper proposes a TLS-and UAV-based method for monitoring landslide deformation in high mountain valleys using traditional real-time kinematics(RTK)-based control points(RCPs),low-precision TLS-based control points(TCPs)and assumed control points(ACPs)to achieve high-precision surface deformation analysis under obstructed vision and impassable conditions.The effects of GCP accuracy,GCP quantity and automatic tie point(ATP)quantity on the accuracy of UAV modeling and surface deformation analysis were comprehensively analyzed.The results show that,the proposed method allows for the monitoring accuracy of landslides to exceed the accuracy of the GCPs themselves by adding additional low-accuracy GCPs.The proposed method was implemented for monitoring the Xinhua landslide in Baoxing County,China,and was validated against data from multiple sources.

APPLICATION OF FUZZY OPTIMIZATION MODEL IN ECOLOGICAL SECURITY PRE-WARNING

Ecological security is a vital problem that people all over the world today have to face and solve, and the situation of ecological security is getting more and more severe and has begun to impede heavily the sustainable development of social economy. Ecological environment pre-warning has become a hotspot for the modern environment science. This paper introduces the theories of ecological security pre-warning and tries to constitute a pre-warning model of ecological security. In terms of pressure-state-response model, the pre-warning guide line of ecological security is constructed while the pre-warning degree judging model of ecological security is established based on fuzzy optimization. As a case, the model is used to assess the present condition pre-warning of the ecological security of Anhui Province. The result is in correspondence with the real condition: the ecological security situations of 8 cities are dangerous and 9 cities are secure. The result shows that this model is scientific and effective for regional ecological security pre-warning.

Underground ground control monitoring and interpretation,and numerical modeling, and shield capacity design

Mine or longwall panel layout is a 3D structure with highly non-uniform stress distribution. Recognition of such fact will facilitate underground problem identification/investigation and solving by numerical modeling through proper model construction. Due to its versatility, numerical modeling is the most popular method for ground control design and problem solving. However numerical modeling results require highly experienced professionals to interpret its validity/applicability to actual mining operations due to complicated mining and geological conditions. Underground ground control monitoring is routinely performed to predict roof behavior such as weighting and weighting interval without matching observation of face mining condition while the mining pressures are being monitored, resulting in unrealistic interpretation of the obtained data on mining pressure. The importance of ground control pressure monitoring and simultaneous observation of mining and geological conditions is illustrated by an example of shield leg pressure monitoring and interpretation in an U.S. longwall coal mine: it was found that the roof strata act like a plate, not an individual block of the size of a shield dimension, as commonly assumed by all researchers and shield capacity is not a fixed property for a longwall panel or a mine or a coal seam. A new mechanism on the interaction between shield's hydraulic leg pressure and roof strata for shield loading is proposed.

Seepage safety monitoring model for an earth rock dam under influence of high-impact typhoons based on particle swarm optimization algorithm

Extreme hydrological events induced by typhoons in reservoir areas have presented severe challenges to the safe operation of hydraulic structures. Based on analysis of the seepage characteristics of an earth rock dam, a novel seepage safety monitoring model was constructed in this study. The nonlinear influence processes of the antecedent reservoir water level and rainfall were assumed to follow normal distributions. The particle swarm optimization （PSO） algorithm was used to optimize the model parameters so as to raise the fitting accuracy. In addition, a mutation factor was introduced to simulate the sudden increase in the piezometric level induced by short-duration heavy rainfall and the possible historical extreme reservoir water level during a typhoon. In order to verify the efficacy of this model, the earth rock dam of the Siminghu Reservoir was used as an example. The piezometric level at the SW1-2 measuring point during Typhoon Fitow in 2013 was fitted with the present model, and a corresponding theoretical expression was established. Comparison of fitting results of the piezometric level obtained from the present statistical model and traditional statistical model with monitored values during the typhoon shows that the present model has a higher fitting accuracy and can simulate the uprush feature of the seepage pressure during the typhoon perfectly.

Monitoring models for base flow effect and daily variation of dam seepage elements considering time lag effect

Affected by external environmental factors and evolution of dam performance, dam seepage behavior shows nonlinear time-varying characteristics. In this study, to predict and evaluate the long-term development trend and short-term fluctuation of the dam seepage behavior, two monitoring models were developed, one for the base flow effect and one for daily variation of dam seepage elements. In the first model, to avoid the influence of the time lag effect on the evaluation of seepage variation with the time effect component of seepage elements, the base values of the seepage element and the reservoir water level were extracted using the wavelet multi-resolution analysis method, and the time effect component was separated by the established base flow effect monitoring model. For the development of the daily variation monitoring model for dam seepage elements, all the previous factors, of which the measured time series prior to the dam seepage element monitoring time may have certain influence on the monitored results, were considered. Those factors that were positively correlated with the analyzed seepage element were initially considered to be the support vector machine(SVM) model input factors, and then the SVM kernel function-based sensitivity analysis was performed to optimize the input factor set and establish the optimized daily variation SVM model. The efficiency and rationality of the two models were verified by case studies of the water level of two piezometric tubes buried under the slope of a concrete gravity dam.Sensitivity analysis of the optimized SVM model shows that the influences of the daily variation of the upstream reservoir water level and rainfall on the daily variation of piezometric tube water level are processes subject to normal distribution.

Cavity 3D modeling and correlative techniques based on cavity monitoring

According to the mining method for Dongguashan Copper Mine and Tongkeng Mine in China, and with the help of the cavity monitoring system（CMS） and mining software Surpac, the 3D cavity models were established exactly. A series of correlative techniques for calculating stope over-excavation and under-excavation, stope dilution and ore loss rates, and the blasting design of the pillar with complicated irregular boundaries were developed. These techniques were applied in Dongguashan Copper Mine and Tongkeng Mine successfully. Using these techniques, the dilution rates of stopes 52-2^#, 52-6^#, 52-8^#and 52-10^# of Dongguashan Copper Mine are calculated to be 2.12%, 8.46%, 12-67% and 10.68%, respectively, and the ore loss rates of stopes 52-6^# and 5-8^# are 4.41% and 3.70%, severally. Furthermore, according to the design accomplished by the technique for a pillar of Tongkeng Mine with irregular boundary, the volume, total length of boreholes and the dynamite quantity of the pillar are computed to be 1.2 ×10^4 m^3, 2.98 km and 10.97 t, correspondingly.

Characteristics analyzing and parametric modeling of the arc sound in CO_2 GMAW for on-line quality monitoring

For on-line monitoring of welding quality, the characteristics of the arc sound signals in short circuit CO2 GMAW were analyzed in the time and frequency domains. The arc sound presents a series of ringing-like oscillations that occur at the end of short circuit i. e. the moment of arc re-ignition, and distributes mainly in the frequency band below 10 kHz. A concept of the arc tone channel and its equivalent electrical model were suggested, which is considered a time-dependent distributed parametric system of which the transmission properties depend upon the geometric and physical characteristics of the arc and surroundings, and is excited by the sound source results from the change of arc energy so that results in arc sound. The linear prediction coding （ LPC ） model is an estimation of the tone channel. The radial basis function （ RBF ） neural networks were built for on-line pattern recognition of the gas-lack in welding, in which the input vectors were formed with the LPC coefficients. The test results proved that the LPC model of arc sound and the RBF networks are feasible in on-line quality monitoring.

Distributed Fiber Optic Monitoring and Stability Analysis of a Model Slope under Surcharge Loading

In the discipline of geotechnical engineering, fiber optic sensor based distributed monitoring has played an increasingly important role over the past few decades. Compared with conventional sensors, fiber optic sensors have a variety of exclusive advantages, such as smaller size, higher precision, and better corrosion resistance. These innovative monitoring technologies have been successfully applied for performance monitoring of geo-structures and early warning of potential geo- hazards around the world. In order to investigate their ability to monitor slope stability problems, a medium-sized model of soil nailed slope has been constructed in laboratory. The fully distributed Brillouin optical time-domain analysis （BOTDA） sensing technology was employed to measure the horizontal strain distributions inside the model slope. During model construction, a specially designed strain sensing fiber was buried in the soil mass. Afterward, the surcharge loading was applied on the slope crest in stages using hydraulic jacks and a reaction frame. During testing, an NBX-6o5o BOTDA sensing interrogator was used to collect the fiber optic sensing data. The test results have been analyzed in detail, which shows that the fiber optic sensors can capture the progressive deformation and failure pattern of the model slope. The limit equilibrium analyses were also conducted to obtain the factors ofsafety of the slope under different surface loadings. It is found that the characteristic maximum strains can reflect the stability of the model slope and an empirical relationship was obtained, This study verified the effectiveness of the distributed BOTDA sensing technology in performance monitoring of slope.

Current Model Analysis of South China Sea Based on Empirical Orthogonal Function(EOF) Decomposition and Prototype Monitoring Data

Environmental load is the primary factor in the design of offshore engineering structures and ocean current is the principal environmental load that causes underwater structural failure. In computational analysis, the calculation of current load is mainly based on the current profile. The current profile model, which is based on a structural failure criterion, is conducive to decreasing the uncertainty of the current load. In this study, we used prototype monitoring data and the empirical orthogonal function(EOF) method to investigate the current profile in the South China Sea and its correlation with the design of underwater structural strength and the dynamic design of fatigue. The underwater structural strength design takes into account the size of the structure and the service water depth. We propose profiles for the overall and local designs using the inverse first-order reliability method(IFORM). We extracted the characteristic profile current(CPC) of the monitored sea area to solve dynamic design problems such as vortex-induced vibration(VIV). We used random sampling to verify the feasibility of using the EOF method to calculate the CPC from the current data and identified the main problems associated with using the CPC, which deserve close attention in VIV design. Our research conclusions provide direct references for determining current load in this sea area. This analysis method can also be used in the analysis of other sea areas or field variables.

Construction of multi-factor identification model for real-time monitoring and early warning of mine water inrush

As a new technical means that can detect abnormal signs of water inrush in advance and give an early warning,the automatic monitoring and early warning of water inrush in mines has been widely valued in recent years.Due to the many factors affecting water inrush and the complicated water inrush mechanism,many factors close to water inrush may have precursory abnormal changes.At present,the existing monitoring and early warning system mainly uses a few monitoring indicators such as groundwater level,water influx,and temperature,and performs water inrush early warning through the abnormal change of a single factor.However,there are relatively few multi-factor comprehensive early warning identification models.Based on the analysis of the abnormal changes of precursor factors in multiple water inrush cases,11 measurable and effective indicators including groundwater flow field,hydrochemical field and temperature field are proposed.Finally,taking Hengyuan coal mine as an example,6 indicators with long-term monitoring data sequences were selected to establish a single-index hierarchical early-warning recognition model,a multi-factor linear recognition model,and a comprehensive intelligent early-warning recognition model.The results show that the correct rate of early warning can reach 95.2%.

The AVO Effect of Formation Pressure on Time-Lapse Seismic Monitoring in Marine Carbon Dioxide Storage

The phase change of CO_(2) has a significant bearing on the siting, injection, and monitoring of storage. The phase state of CO_(2) is closely related to pressure. In the process of seismic exploration, the information of formation pressure can be response in the seismic data. Therefore, it is possible to monitor the formation pressure using time-lapse seismic method. Apart from formation pressure, the information of porosity and CO_(2) saturation can be reflected in the seismic data. Here, based on the actual situation of the work area, a rockphysical model is proposed to address the feasibility of time-lapse seismic monitoring during CO_(2) storage in the anisotropic formation. The model takes into account the formation pressure, variety minerals composition, fracture, fluid inhomogeneous distribution, and anisotropy caused by horizontal layering of rock layers(or oriented alignment of minerals). From the proposed rockphysical model and the well-logging, cores and geological data at the target layer, the variation of P-wave and S-wave velocity with formation pressure after CO_(2) injection is calculated. And so are the effects of porosity and CO_(2) saturation. Finally, from anisotropic exact reflection coefficient equation, the reflection coefficients under different formation pressures are calculated. It is proved that the reflection coefficient varies with pressure. Compared with CO_(2) saturation, the pressure has a greater effect on the reflection coefficient. Through the convolution model, the seismic record is calculated. The seismic record shows the difference with different formation pressure. At present, in the marine CO_(2) sequestration monitoring domain, there is no study involving the effect of formation pressure changes on seismic records in seafloor anisotropic formation. This study can provide a basis for the inversion of reservoir parameters in anisotropic seafloor CO_(2) reservoirs.

Finite element model updating of existing steel bridge based on structural health monitoring

Based on the physical meaning of sensitivity,a new finite element(FE) model updating method was proposed. In this method,a three-dimensional FE model of the Nanjing Yangtze River Bridge(NYRB) with ANSYS program was established and updated by modifying some design parameters. To further validate the updated FE model,the analytical stress-time histories responses of main members induced by a moving train were compared with the measured ones. The results show that the relative error of maximum stress is 2.49% and the minimum relative coefficient of analytical stress-time histories responses is 0.793. The updated model has a good agreement between the calculated data and the tested data,and provides a current baseline FE model for long-term health monitoring and condition assessment of the NYRB. At the same time,the model is validated by stress-time histories responses to be feasible and practical for railway steel bridge model updating.